Sign Language (Prerecorded)
Prerecorded audio that runs synchronized with something else needs a sign language interpretation, and that sits on top of captions rather than replacing them. Video with sound is the usual case. A narrated slide deck is the one people miss, because there is no video track and the rule still reaches it. A podcast with nothing running alongside it falls outside. Both routes W3C documents run synchronized with the original, either an interpreter composited into the video itself or a second video stream the player lines up beside it.
Why it matters
A sign language is a language in its own right, unrelated to the spoken language of the country it is used in, which is why ASL, BSL and Auslan are three different languages rather than three accents. Black American Sign Language is distinct from ASL too. For somebody who signs, written text is often a second language, and where it is, captions are a translation they read at second-language speed while the video keeps running at its own. Signing also carries intonation and feeling that captions flatten out. And not every deaf person signs, especially people who lost their hearing later in life, which is exactly why this rule adds to captions instead of standing in for them.
Who this rule protects
Deaf people whose first language is a sign language are the ones this protects, and W3C notes plainly that some of them have limited reading ability, so captions do not reach them. Some viewers want captions and signing on screen at the same time.
How to check it yourself
- List the prerecorded recordings whose sound runs synchronized with something else, meaning video with speech and narrated slides alike. A recording that is only sound, with nothing running beside it, is out of scope.
- Check whether each one offers a sign language track or an interpreted version.
- Confirm the interpretation runs synchronized with the video. A signed summary published on its own is not a documented route to meeting this.
- Check that the signed and unsigned versions are both easy to find from the player itself.
Failures we see most often
- Captions get treated as a substitute, when for a first-language signer they are a translation into a second language.
- An interpretation is published as a separate clip that never lines up with the original video.
- The signer is framed too tightly to show the full signing space, which runs from below the waist to above the head and an elbow width to each side.
- A signing avatar stands in for a human interpreter.
Who this one is for
Read from this rule's own note above, so the grouping and the note cannot disagree.
- Deaf and hard of hearingpeople who cannot hear the audio
How to fix it
- Commission interpretation for your highest-value evergreen video first, and pick the sign language your primary audience actually uses.
- Deliver it as a synchronized track, either composited into the video or as a second stream the player overlays.
- Shoot it properly. Solid background, clothing that contrasts with the signer's skin tone, good light, and the signer placed clear of anything important on screen.
- Where you cannot cover everything yet, say so in your accessibility statement rather than letting the gap go unmentioned.
Step-by-step fix guides (4)
Passes vs. fails
Passes
The same video adds a BSL interpreter in a picture-in-picture track, with a labelled toggle sitting right under the player.
Fails
An hour of onboarding video ships with accurate captions and nothing else, though the product is sold to deaf customers.
In audits and lawsuits
This is Level AAA, so it sits outside our audit scope and outside every law we track. W3C says the same thing in its own words. Sign language is not required by most web accessibility policies. It is also the criterion that shows most clearly why AAA should not be demanded across a whole site, because a publisher shipping video every day cannot interpret all of it. One warning if you go looking for a shortcut. W3C's position on signing avatars is that they are not good enough to be adequate.